To further test the hypothesis that urinary dithiocarbamate excretion can be specifically attributed to ingestion of cruciferous vegetables, we conducted an eighteen day inpatient study in four healthy volunteers (30-75 years old; 2 black, 2 white; 4 males). There eas a two day washout period, followed by four days of daily dosing with 250 g of a n0n-cruciferous vegetable (green beans, corn, tomatoes, or carrots), then by a 12 day period during which 250 g of a cruciferous vegatable (kale, broccoli, green cabbage, or turnip root) was administered every three days. Within these two categories, the vegatables were administered in random order, such that no two volunteers received the same vegetable on any day. All eight vegetables were chopped (except corn), heated in a microwave to destroy any myrosinase activity, and stored at -20 C. Aliquots were analyzed for glucosinolate and isothiocyanate content, and for Phase 2 enzyme inducer activity. The four non-cruciferous vegetables had no detectable glucinolates, isothiocyanantes, or inducer activity. The crucifers contained 0.3 to 0.8 mol/ g fresh weight non-indole isothiocyanate plus glucosinolate. Only in kale did free isothiocyanates constitute a substantial proportion (43%) of the total activity. Glucosinolate profiles revealesd that all four vegetables had rather complex compositions. In broccoli, glucoraphanin predominated, along with the indole glucosinolates, 4-hydroxyglucobrassicin, glucobrassicin, and neoglucobrassicin. In cabbage, glucoraphanin, sinigrin and glucobrassicin were the primary glucosinolates. In kale, glucoraphanin was the major glucosinolate, and in turnip roots, progoitrin, sinigrin, and several indole glucosinolates were identified.